This model is a fine-tuned version of
Helsinki-NLP/opus-mt-ko-en on the kde4 dataset.
It achieves the following results on the evaluation set:
You can use this model directly with a pipeline for translation language modeling:
1>>> from transformers import pipeline
2>>> translator = pipeline('translation',model='chunwoolee0/kd4_opus-mt-ko-e')
3>>> translator("점심 식사 후에 산책가자.")
4
5[{'translation_text': "Let's go for a walk after noon."}]
6
7>>> translator("이 강좌는 허깅페이스가 만든 거야.")
8[{'translation_text': 'This is a course by Huggingspace.'}]
9
10>>> translator("오늘은 늦게 일어났다.")
11[{'translation_text': "I'm up late today."}]
Step Training Loss
500 1.858500
1000 1.781400
1500 1.715200
2000 1.678100
2500 1.546600
3000 1.488700
3500 1.503500
4000 1.455100
4500 1.419100
5000 1.393400
5500 1.357100
6000 1.339400
TrainOutput(global_step=6474, training_loss=1.532715692246148,
metrics={'train_runtime': 1035.7775, 'train_samples_per_second': 199.957,
'train_steps_per_second': 6.25, 'total_flos': 2551308264603648.0,
'train_loss': 1.532715692246148, 'epoch': 3.0})
